📘 Available Parameters (--help)

An overview table of all arguments that OmniOpt2 accepts via the Command line. Same data as in ./omniopt --help.

Parameter	Description	Type	Action	Default Value
Ungrouped
`--config_yaml`	YAML configuration file	`str`
`--config_toml`	TOML configuration file	`str`
`--config_json`	JSON configuration file	`str`
Required arguments - These options have to be set
`--num_random_steps`	Number of random (SOBOL) steps to start with	`int`		`20`
`--max_eval`	Maximum number of evaluations	`int`
`--run_program`	A program that should be run. Use, for example, $x for the parameter named x	`str`	`append`
`--experiment_name`	Name of the experiment	`str`
`--mem_gb`	Amount of RAM for each worker in GB (default: 1GB)	`int`		`1`
Required arguments that allow a choice - Of these arguments, one has to be set to continue
`--parameter`	Experiment parameters in the formats (options in round brackets are optional): <NAME> range <LOWER BOUND> <UPPER BOUND> (<INT, FLOAT>, log_scale: True/False, default: false>) -- OR -- <NAME> fixed <VALUE> -- OR -- <NAME> choice <Comma-separated list of values>		`append`
`--continue_previous_job`	Continue from a previous checkpoint, use run-dir as argument	`str`
Optional - These options are optional
`--experiment_constraints`	Constraints for parameters. Example: x + y <= 2.0. Convert them to base64	`str`	`append`
`--run_dir`	Directory, in which runs should be saved. Default: runs	`str`		`runs`
`--seed`	Seed for random number generator	`int`
`--decimalrounding`	Number of decimal places for rounding	`int`		`4`
`--verbose_tqdm`	Show verbose TQDM messages	`bool`	`store_true`	`False`
`--model`	Use special models for nonrandom steps. Valid models are: "SOBOL", "FACTORIAL", "SAASBO", "BOTORCH_MODULAR", "UNIFORM", "BO_MIXED", "RANDOMFOREST", "EXTERNAL_GENERATOR", "PSEUDORANDOM", "TPE"	`str`
`--gridsearch`	Enable gridsearch	`bool`	`store_true`	`False`
`--occ`	Use optimization with combined criteria (OCC)	`bool`	`store_true`	`False`
`--show_sixel_scatter`	Show sixel graphics of scatter plots in the end	`bool`	`store_true`	`False`
`--show_sixel_general`	Show sixel graphics of general plots in the end	`bool`	`store_true`	`False`
`--show_sixel_trial_index_result`	Show sixel graphics of trial index in the end	`bool`	`store_true`	`False`
`--follow`	Automatically follow log file of sbatch	`bool`	`store_true`	`False`
`--send_anonymized_usage_stats`	Send anonymized usage stats	`bool`	`store_true`	`False`
`--ui_url`	Site from which the OO-run was called	`str`
`--root_venv_dir`	Where to install your modules to ($root_venv_dir/.omniax_..., default: $HOME)	`str`
`--exclude`	A comma separated list of values of excluded nodes (taurusi8009,taurusi8010)	`str`
`--main_process_gb`	Amount of RAM for the main process in GB (default: 8GB)	`int`		`8`
`--max_nr_of_zero_results`	Max. nr of successive zero results by the generator before the search space is seen as exhausted	`int`		`10`
`--abbreviate_job_names`	Abbreviate pending job names (r = running, p = pending, u = unknown, c = cancelling)	`bool`	`store_true`	`False`
`--orchestrator_file`	An orchestrator file	`str`
`--checkout_to_latest_tested_version`	Automatically checkout to latest version that was tested in the CI pipeline	`bool`	`store_true`	`False`
`--live_share`	Automatically live-share the current optimization run automatically	`bool`	`store_true`	`False`
`--disable_tqdm`	Disables the TQDM progress bar	`bool`	`store_true`	`False`
`--disable_previous_job_constraint`	For continued jobs: Disable getting the constraint of the previous job that is about to be continued	`bool`	`store_true`	`False`
`--workdir`	Working directory	`str`
`--occ_type`	Optimization-with-combined-criteria-type (valid types are "geometric", "euclid", "signed_harmonic", "signed_minkowski", "weighted_euclid", "composite")	`str`		`euclid`
`--result_names`	Name of hyperparameters. Example --result_names result1=max result2=min result3. Default: RESULT=min
`--minkowski_p`	Minkowski order of distance (default: 2), needs to be larger than 0	`float`		`2`
`--signed_weighted_euclidean_weights`	A comma-seperated list of values for the signed weighted euclidean distance. Needs to be equal to the number of results. Else, default will be 1	`str`
`--generation_strategy`	A string containing the generation_strategy. Example: SOBOL=10,BOTORCH_MODULAR=10,SOBOL=10. Cannot use --model EXTERNAL_GENERATOR or PSEUDORANDOM	`str`
`--generate_all_jobs_at_once`	Generate all jobs at once rather than to create them and start them as soon as possible	`bool`	`store_true`	`False`
`--revert_to_random_when_seemingly_exhausted`	Generate random steps instead of systematic steps when the search space is (seemingly) exhausted	`bool`	`store_true`	`False`
`--load_data_from_existing_jobs`	List of job data to load from existing jobs	`str`
`--n_estimators_randomforest`	The number of trees in the forest for RANDOMFOREST (default: 100)	`int`		`100`
`--max_attempts_for_generation`	Max. number of attempts for generating sets of new points (default: 20)	`int`		`20`
`--external_generator`	Programm call for an external generator	`str`
`--username`	A username for live share	`str`
`--max_failed_jobs`	Maximum number of failed jobs before the search is cancelled. Is defaulted to the value of --max_eval	`int`
`--num_cpus_main_job`	Number of CPUs for the main job	`int`
`--calculate_pareto_front_of_job`	This can be used to calculate a pareto-front for a multi-objective job that previously has results, but has been cancelled, and has no pareto-front (yet)	`str`
`--show_generate_time_table`	Generate a table at the end, showing how much time was spent trying to generate new points	`bool`	`store_true`	`False`
`--force_choice_for_ranges`	Force float ranges to be converted to choice	`bool`	`store_true`	`False`
Speed - These options are for speeding up the Process of Starting Processes or Generating new Points
`--dont_warm_start_refitting`	Do not keep Model weights, thus, refit for every generator (may be more accurate, but slower)	`bool`	`store_true`	`False`
`--refit_on_cv`	Refit on Cross-Validation (helps in accuracy, but makes generating new points slower)	`bool`	`store_true`	`False`
`--fit_out_of_design`	Ignore data points outside of the design while creating new points	`bool`	`store_true`	`False`
`--fit_abandoned`	Do not ignore abandoned data points while creating new points	`bool`	`store_true`	`False`
`--dont_jit_compile`	Disable JIT-compiling the model	`bool`	`store_true`	`False`
`--num_restarts`	num_restarts option for optimizer_options	`int`		`20`
`--raw_samples`	raw_samples option for optimizer_options	`int`		`1024`
`--max_num_of_parallel_sruns`	Maximal number of parallel sruns	`int`		`16`
`--no_transform_inputs`	Disable input transformations	`bool`	`store_true`	`False`
`--no_normalize_y`	Disable target normalization	`bool`	`store_true`	`False`
`--transforms`	Enable input/target transformations (choose one or both: Cont_X_trans, Cont_X_trans_Y_trans)
SLURM - Parameters related to SLURM
`--num_parallel_jobs`	Number of parallel SLURM jobs (default: 20)	`int`		`20`
`--worker_timeout`	Timeout for SLURM jobs (i.e. for each single point to be optimized)	`int`		`30`
`--slurm_use_srun`	Using srun instead of sbatch	`bool`	`store_true`	`False`
`--time`	Time for the main job	`str`
`--partition`	Partition to be run on	`str`
`--reservation`	Reservation	`str`
`--force_local_execution`	Forces local execution even when SLURM is available	`bool`	`store_true`	`False`
`--slurm_signal_delay_s`	When the workers end, they get a signal so your program can react to it. Default is 0, but set it to any number of seconds you wish your program to be able to react to USR1	`int`		`0`
`--nodes_per_job`	Number of nodes per job due to the new alpha restriction	`int`		`1`
`--cpus_per_task`	CPUs per task	`int`		`1`
`--account`	Account to be used for SLURM	`str`
`--gpus`	Number of GPUs per worker	`int`		`0`
Installing - Parameters related to installing
`--run_mode`	Either local or docker	`str`		`local`
Debug - These options are mainly useful for debugging
`--verbose`	Verbose logging	`bool`	`store_true`	`False`
`--verbose_break_run_search_table`	Verbose logging for break_run_search	`bool`	`store_true`	`False`
`--debug`	Enable debugging	`bool`	`store_true`	`False`
`--flame_graph`	Enable flame-graphing. Makes everything slower, but creates a flame graph	`bool`	`store_true`	`False`
`--no_sleep`	Disables sleeping for fast job generation (not to be used on HPC)	`bool`	`store_true`	`False`
`--tests`	Run simple internal tests	`bool`	`store_true`	`False`
`--show_worker_percentage_table_at_end`	Show a table of percentage of usage of max worker over time	`bool`	`store_true`	`False`
`--auto_exclude_defective_hosts`	Run a Test if you can allocate a GPU on each node and if not, exclude it since the GPU driver seems to be broken somehow	`bool`	`store_true`	`False`
`--run_tests_that_fail_on_taurus`	Run tests on Taurus that usually fail	`bool`	`store_true`	`False`
`--raise_in_eval`	Raise a signal in eval (only useful for debugging and testing)	`bool`	`store_true`	`False`
`--show_ram_every_n_seconds`	Show RAM usage every n seconds (0 = disabled)	`int`		`0`
`--show_generation_and_submission_sixel`	Show sixel plots for generation and submission times	`bool`	`store_true`	`False`